In Vitro Efficacy of Antiviral Agents against Omicron Subvariant BA.4.6

To the Editor: As of September 2022, the BA.5 subvariant of the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) B.1.1.529 (omicron) variant has become dominant in most countries around the world. However, the prevalence of the BA.4.6 subvariant is increasing in the United States.¹ BA.4.6 and BA.5 share the same amino acid substitutions in the receptor-binding domain of the spike protein, which is the major target for vaccines and therapeutic monoclonal antibodies against SARS-CoV-2. BA.4.6 also has an additional mutation that is not found in BA.5 (i.e., R346T),² a finding that arouses concern that the effectiveness of current vaccines and therapeutic monoclonal antibodies against this subvariant will be greatly decreased.

Accordingly, we assessed the effectiveness of several therapeutic monoclonal antibodies that have been authorized for the treatment of coronavirus disease 2019 (Covid-19), individually and in combination, against two omicron BA.4.6 isolates — hCoV-19/USA/WI-UW-12757/2022 (UW-12757) and hCoV-19/USA/WI-UW-12767/2022 (UW-12767). Both isolates had been obtained from patients with Covid-19. The spike protein of these two isolates contained two additional amino acid changes (R346T and N658S) as compared with a BA.5 isolate (hCoV-19/Japan/TY41-702/2022) (Fig. S1A in the Supplementary Appendix, available with the full text of this letter at NEJM.org). In addition, one of the isolates (UW-12757) also had an N487D mutation in its receptor-binding domain.

We used a live-virus 50% focus reduction neutralization test (FRNT₅₀) to determine neutralization titers of monoclonal antibodies, including REGN10987 (marketed as imdevimab) and REGN10933 (marketed as casirivimab). REGN10987 retained some neutralizing activity against the two BA.4.6 isolates (Figure 1A and 1B), but REGN10933 did not retain such activity. REGN10987 in combination with REGN10933 (imdevimab–casirivimab) neutralized both BA.4.6 isolates; however, as compared with the ancestral strain, the effectiveness of this combination was lower by a factor of 52.9 against UW-12757 and by a factor of 87.3 against UW-12767. The monoclonal antibodies COV2-2196 (marketed as tixagevimab) and COV2-2130 (marketed as cilgavimab), individually and in combination, had reduced activity against the BA.4.6 isolates, as did S309, the precursor of sotrovimab. In contrast, LYCoV1404 (marketed as bebtelovimab) efficiently inhibited both UW-12757 and UW-12767 with a very low FRNT₅₀ value (3.80 ng per milliliter and 2.26 ng per milliliter, respectively), results that were similar to those for the ancestral strain.

Figure 1. Antiviral Efficacy and Antibody Response in Vitro against Omicron Subvariants.

Shown is the neutralizing activity (Panel A) and efficacy (Panel B) of monoclonal antibodies and the inhibitory activity (Panel C) and efficacy (Panel D) of antiviral drugs against omicron subvariants. GS-441524 (the main metabolite of remdesivir) and EIDD-1931 (the active form of molnupiravir) are RNA-dependent RNA polymerase inhibitors. PF-07321332 (nirmatrelvir) is an Mpro inhibitor. Also shown are the neutralizing activity of plasma obtained from patients who had received three doses of the BNT162b2 vaccine (Panel E, left) and the neutralizing activity of plasma obtained from patients who had been infected with the omicron BA.2 subvariant after receiving either three doses of the BNT162b2 vaccine or two doses of the mRNA-1273 vaccine and one dose of the BNT162b2 vaccine (Panel E, right). Detailed information about the participants is provided in Tables S3 and S4. IC₅₀ denotes 50% inhibitory concentration, and FRNT₅₀ 50% focus reduction neutralization test.

The Food and Drug Administration has approved the use of remdesivir (an RNA-dependent RNA polymerase [RdRp] inhibitor) for the treatment of Covid-19 and has issued Emergency Use Authorizations for two other antiviral drugs: molnupiravir (an RdRp inhibitor) and nirmatrelvir (a main protease inhibitor of SARS-CoV-2). We therefore tested the efficacy of these antiviral drugs against BA.4.6 by determining their in vitro 50% inhibitory concentration (IC₅₀) values against this variant. Of note, these two BA.4.6 isolates had P314L and P3395H mutations in the RdRp and main protease, respectively (Fig. S1B). The two BA.4.6 isolates had susceptibilities to the three compounds that were similar to the susceptibility of the ancestral strain: for UW-12757, the IC₅₀ value was higher by a factor of 1.6 with remdesivir, by a factor of 5.7 with molnupiravir, and by a factor of 4.1 with nirmatrelvir; for UW-12767, the IC₅₀ values were higher by a factor of 0.4, 1.8, and 1.2, respectively (Figure 1C and 1D). The clinical C_max (i.e., the highest concentration of a drug in the blood) of the therapeutic agents is shown in Tables S1 and S2.

The neutralizing activity of plasma obtained from patients who had recovered from Covid-19 and from recipients of vaccines was lower against BA.4.6, BA.2, and BA.5 than it was against the ancestral strain. This reduction in neutralizing titers was larger for BA.4.6 and BA.5 than for BA.2 (Figure 1E and Supplementary Results and Tables S3 and S4).

Our data suggest that remdesivir, molnupiravir, and nirmatrelvir and the monoclonal antibodies bebtelovimab and imdevimab retain effectiveness against BA.4.6 in vitro (see the Supplementary Discussion). Our findings also indicate that monoclonal antibodies casirivimab, sotrovimab, tixagevimab, and cilgavimab may not be effective against BA.4.6.

This letter was published on November 16, 2022, at NEJM.org.